@rawbrawb Your tag-edit of Ohms-Law, there are a number of grammatical errors. I'll reject it since it does not allow me to edit and accept, ok?
A well know relationship that describes the relationship between Voltage and current through a devices resistance. V= IR which means that voltage across the device is equal to the current through the device multiplied by the resistance.
Same issue here: Also know as SOC it is a for of Silicon chip which has a high degree of integration - the so called System on a chip. Called principally because it incorporates devices like eg. a uProcessor and a modem and a memory controller that might of at one time been separate chips on a PCB.
@jippie Differential signals need not be ground referenced at all. Voltage across a current shunt can be arbitrary. So can a signal from a microphone (though you could force a ground reference on those). Often professional mics will use the stage equipment ground, which need not be your circuit's ground.
@jippie Yes you can use a diffamp with a floating signal as long as the float does not go beyond the allowed common-mode signal level for the amp.
@jippie Which is why (mic example again) sometimes a bit of bias-to-ground is added through a multi-megohm resistor - insufficient to drag reference to ground, and high enough to not couple in hum between grounds.
Another advantage is even though something might be ground referenced think of something where one side is 11V to ground and other is 12V. With differential you could use a 1.2V reference and get lots of resolution instead of a divider to get it into range.
although thinking further maybe Anindo knows, I'm not sure how most microcontrollers go about clamping in that case, I was mainly thinking about external ADCs
Sounds like a good one to me, although I suppose the only thing is that lots of different micros will have different differential input topologies (or none at all) so it could be a bit open ended without a particular part or series of micros.
@jippie, looking at an ATMega128 for example it looks like in differential mode the input still has to be 0 - Vcc, I thought that might be the case, some external ADCs aren't that way. So you'd still get the advantage (especially if using gain) of the better resolution but only over that range.
@PeterJ Yup, standalone diffamps often go well beyond rails in common-mode. MCU differential inputs actually aren't: They are not truly differential independent of (or at least beyond) supply rails.
@PeterJ Basically the MCU diff inputs are faking it, so to speak. Good for scale expansion, not for the other purposes a Differential Amp is used for.
@PeterJ Yeah, I just avoid it on MCUs. Also, I'm obsessed with various breeds of op-amp anyway, including diff-amps, so I'd rather do a unity-gain on a diffamp and use output to drive a single MCU pin, instead of wasting an additional pin of the MCU.
@AnindoGhosh Thanks, a combination of small edit window,no preview and (mainly) tired brain. Since edited. I had caught at least some of those errors before ... so I wonder what happened. maybe just another tired brain again.
@AnindoGhosh I don't know about external diff amps (wrt beyond the rails) but all ADC's that I've seen in chips are differential but are limited to within the rails for obvious reasons. Transistor Vgs limits and latch-up reasons. However they are truly differential and typically have common mode feedback circuits (internal)
@rawbrawb Yep, there are those which limit common mode (and thus limit their utility to my purposes), but the neat ones are not limited to within-the-rails... and I don't mean the ISOamps alone.
@AnindoGhosh could you send me a part number of one example? If they use magnetics I can understand but a pure Si solution I'd like to know what they are talking about.
This is an interesting one too: analog.com/en/specialty-amplifiers/difference-amplifiers/… - I quote: Unlike most types of amplifiers, difference amplifiers are typically able to measure voltages beyond the supply rails, and are used in applications where large dc or ac common-mode voltages are present.
What I like about Analog is their lead-outs to related product families, e.g. "For higher speed and fully differential amplifiers, please see differential amplifiers. For amplifiers with higher input impedance, please see instrumentation amplifiers." - That's a trick I wish TI would pick up.
@rawbrawb That's one mechanism, there are a couple of others as well. All pre-1980 stuff except the iCoupler stuff, which truly is a wonder. actual physical transformer coils on a silicon substrate? That's totally insane.
Now I'm waiting for Apple to sue Analog Devices for using the letter i in iCoupler.
@rawbrawb While Analog Devices got themselves the patents on magnetic transformers within a chip, TI went the other way, they have piezo coupling transformers on-wafer, to compete with iCoupler.
It makes me think about some of the MEMS stuff I've been working on. The forces you're able to generate at small gaps are truly amazing, via the E-Field.
@AnindoGhosh piezo on chip? must be another layer on top. Si has Piezo resistance but not a piezo-electric effect per se.
@rawbrawb I can well imagine, just extrapolating from the insane capacitance possible in a wafer, with a polyimide dielectric that is a few molecules thick.
@rawbrawb Yes, it's a layer of umm some polymer with piezoelectric characteristics. I admit I may have dozed off briefly during that seminar.
@rawbrawb While SiO is a great discovery, it isn't the only thing that is used as a dielectric on chips. Polyimide for instance is very popular for isolation.
@rawbrawb A picture tour of me napping at seminars would be ... awkward.
@AnindoGhosh well, only in some cases that are very limited. It is used as a top passivation, but Si3N4 is far superior for film quailty. Polyimide is useful in some cases though but limits life.
@rawbrawb That's a good start too :-) If any fab did let me inside, they'd be very upset - I have a crazy lack of brain-to-mouth-filter, and I almost invariably spot something that could be done differently if not provably better.
@rawbrawb I've even spotted optimization possibilities in a rum bottling factory of the world's third-largest selling rum. I was only there on a courtesy visit, my company was designing their web site.
@Dean Have you thought about just posting a question with the schematic and code? Maybe just a cut-down self contained example that tries to read and write a single byte and doesn't do anything else but configure the SPI and do that.
I will do later if im still struggling, there just seems to be so many inconsistencies throughout everything though. Compilers being the biggest problem
OK, just seems like a good and on-topic question if you narrow it down a bit and at least a few more people will read it and it'll hang around a bit longer than it chat.
@Dean Then why not set up an I2C EEPROM explorer on the Arduino to see how things are getting written or not - I have read multiple reports of MSP430 problems reading back from an I2C EEPROM - consensus seems to be some issue with the 3.3 Volt signal levels of the MSP430.
@Dean Another thing, why not try with the IAR compiler first, then get engaged in porting the core to GCC?
@Dean Being bogged down due to compiler religion, I don't understand at all.
@Dean Then have you verified that what you are trying works at all with IAR? If that's the only difference, bypass that difference to validate whether the issue lies elsewhere.
@AnindoGhosh, what about Altavista? I'd forgotten about that until recently when someone mentioned in a comment and see it's still 'somewhat' alive but taken over by Yahoo.
@PeterJ Yes. Yahoo apparently only bought it to shut it down (underscoring my cynical outlook about Yahoo). They used the brand for a while, now it's just Yahoo Search in Altavista sheepskin.
@Dean If you have the launchpad limited free edition of IAR, I think it only supports the launchpad, so most likely not. I usually use CCS, but I can fire up IAR which I have installed on some machine around here, if you need me to.
@AnindoGhosh, I remember now a DEC rep sprouting how good it was in the early years. Better not mention RSTS/E and VAX/VMS otherwise I'll get flagged as an old fart.
@PeterJ It was actually mindblowingly good for its time. That's most likely what triggered Yahoo's hostile buyout. Yahoo has been very emotional in its business behavior of late. Case in point...
@PeterJ A few years ago, Yahoo stock started dropping fast, they were in deep doodoo (still are). Microsoft offered a buy-out that was more lucrative than any other offer they had, they declined because it was Microsoft. Less than a year later, they were courting Microsoft for the same deal, but by then Microsoft decided it was not interesting any more. I can understand that response.
@AnindoGhosh, I remember that too when it was the dominent search engine, WebCrawler was another good one at some stage too. Anyway I might head off soon, just about turn into a pumpkin time locally.
@Rocketmagnet, not 100% sure but don't think you can do that. AFAIK you could award a bounty to one person, then add a new bounty and give it to to a second person but not sure.
Anyway what I do know is I need my beauty sleep, catch you all later.
What is a bounty?
What is the "Featured" tab on the homepage?
How can I search for questions that have a bounty attached?
How do I start a bounty? When can I start a bounty?
How long is the bounty period?
How do I award a bounty?
Can I award a bounty to my own answer?
Can I award a bounty to an ...
Maybe I should start calling myself "blows up LEDs for fun". I just caused a nice explosion on a red 3mm LED, but running just under 1 Amp through it. :-)
For a practical answer to the question, destructive testing of at least one LED, preferably a few, will be required.
Broadly:
LEDs are primarily destroyed by heat, not so much by current. Depending on the internal construction of the LED and its short-term thermal dissipation performance, an ...
@AnindoGhosh You should add the content of your comment (about the LED going POP) into the answer. Some people won't read the comments, but there is much useful information there.
I have a MOSFET circuit with multi-threshold voltages. I am using TSMC MOSFET with 180nm technology. How to specify these voltages in LTSPICE for simulation. Please help.
@AnindoGhosh thanks, I answered as best I can (I don't use LTSPice) so while it's generic it may be still to confusing.
@AnindoGhosh these actions certainly undercut the "keeping it pure" credo that was stated as the reason for down votes in that Meta question. My fave was the down vote of laplace transforms.
that question is a perfect question for @ThePhoton to get more rep.
sorry can't hang around, gotta go see a man about some electric fields , hopefully no electric sheep are there.
@rawbrawb See my comment. If the community administration / mods are incapable of leashing the unwelcoming attitudes, I'm hoping at least some of them can be shamed into being a little more welcoming to newcomers.
@AnindoGhosh one comment on your comment, those aren't ivory towers, that implies some level of mathematical analysis. Thise type of question tend to get run out of the house too. it's the OPAL thing again.
I'm currently writing documentation for a school project. Sadly I couldn't get it working in time, as the I2C emulation on the SPI line refused to work.
I am planning on buying a toroid core from Digi-Key. I want to be sure that this is the right type and that I can achieve 170 uh inductance with 22-18 AWG wire. How can I wind this and what is the formula so I can calculate this myself in the future. If this doesn't work which one can I buy from ...